chazeon

As many have discussed earlier, if you have an accent, macOS transcribing is not so accurate; on top of that, Apple's transcribing also don't work well if you have a lot of technical terms (for example, if you are transcribing a thesis in physics)

the place where zotero and alike become useful is when you start to need a tree-like structure

Why does this lab test only have marketing materials? They are not even installing linux on a supposedly LLM dev machine

Well, when 5.5 first came out, it was kind of OK, but now it's almost noticeably worse. It can be done, but requires a lot of effort, just more and more round, and actually, the Gemini Pro on the web (which should be 3.1 Pro) is actually doing a more stable job.

The thing that I ask it to do is like take X and Y paper into Z paragraph --- a not-so-silly model should think of how information in X and Y are related and how they support the whole article to synthesize this sentence in a way that is coherent to the article, but 5.5 now will just copy the stuff without any reasoning about the relation. Of course, this will cost a lot of tokens and will be obvious if not done. One clear indicator is that in a few rounds you can see the length of the article get bloated to 2-3x undesirably long, which is clearly because it is not analyzing/synthesizing the info.

Even without stats i know it went bad. In the pass two month barely can do any good scientific writing lately, which of course rely on reasoning. It just writing for gods sake. And it show how far we are from AGI.

I remember years ago, Ubuntu, Kubuntu, etc., could all be obtained via CD for free; years later, everything become a limited print. https://web.archive.org/web/20090219122023/https://shipit.ub...

I have always been thinking LUKS was supposed to be enrolled in TPM, so you should not have to enter this key manually; this is just to prevent someone from unplugging the hard drive and reading on another machine. Of course, this depends on one's threat model.

But if you do this, don't you have to enter two passwords each time you wake? One for LUKS, one for the system login?

Isn't this the company that make the LLM become a degenerate when it comes to bioscience?

Mac mini Pro line is doomed, they never made enough of it; skipped M5 Pro, now skipping M6 Pro, it is like 2014-2018 again. Now ordering a custom M4 Pro build take 3 months+ to ship with an increased price.

GPT-5.5 must have serious issues; it is fast, but quality-wise, it is just not good. It read one LaTeX paper (which is not long) and can spell my name wrong. This is GPT-5.5-high.

Most CA households may be, but obviously not everywhere

My understanding of safety concern around Linux is mainly the security, namely, full disk encryption. Full disk encryption avoid an attacker unplug your harddrive and directly read the data.

While it is possible, it is definitely not easy to setup right, particularly if you want hibernation to disk properly setup. There is certain requirement on the disk layout setup, use LVM, setup TPM, setup bootloader parameters, setup hibernation and wake.... any step is wrong you have to use a boot drive to rescue, and it is very hard to fix if you don't have LVM in the first place. For example, Arch Linux's archinstall won't setup this whole suite for you.

This is really nessary if you are going to take the computer outside and it might get lost stolen. You definitely don't want other people to read the content after it was lost. I think this kind of security is default on Windows / Mac / iOS / Android right now already, but Linux is still so hard.

Well that’s a very misleading thing. If the US immigration policy wasn’t this hostile to populous countries, more Chinese will want to stay.

Nvidia's private driver seems to deliver 4k@120Hz just fine.

Well the seemingly cheap comes with significantly degraded performance, particular for agentic use. Have you tried replacing Claude Code with some locally deployed model, say, on 4090 or 5090? I have. It is not usable.

Seem images on GitHub web also not showing

The one thing I wish it has is 3.5mm audio jack. Both Xbox and SONY's dualsense controller have this. But SONY don't support audio via Bluetooth. The Xbox one need a USB adapter but its build is not as good as SONY's. SONY don't have a USB adapter. Given Steam controller is already using an USB puck, it should be able to support it.

Gemini is the only model that can provide consistent solution to theoretical physics problems and output it into LaTeX document.

you can just buy a nanokvm

Aliexpress has them

Infuse is good, but it does not feel so well-polished for the desktop, for example, some windows for pop-up could have been a real window, but were a pop-up that blocks the main player.

I want to note that: long prompts are good only if the model is optimized for it. I have tried to swap the underlying model for Claude Code. Most local models, even those claimed to work with long context and tool use, don't work well when instruction becomes too long. This has become an issue for tool use, where tool use works well in small ChatBot-type conversation demos, but when Claude's code-level prompt length increases, it just fails, either forgetting what tools are there, forgetting to use them, or returning in the wrong formats. Only the model by OpenAI, Google's Gemini, kind of works, but not as well as Anthropic's own models. Besides they feel much slower.

I do japanese transcription + gemini translations. It’s worse than fansub, but its much much better than nothing. First thing that could struggle is actually the vad, then is special names and places, prompting can help but not always. Finally it’s uniformity (or style). I still feel that I can’t control the punctuation well.

Recently, I visited the Pennsylvania Railroad Museum and was fascinated to learn that when steel railcars were first introduced—despite being far safer than their wooden predecessors, which could easily be crushed—many people feared they might attract lightning. It's such a good analogue to our movement into AI reality.

Boards are low tech and low profit, does American company and workers even want to do it?

Actually recent email innovation I enjoyed is Mimestream, the macOS native client for Gmail. Apple’s smart inbox is half baked but better than nothing. Cloudflare now also has a pretty good email forwarding service.

The thing is, the publisher are not obligated to show all recent post, some may show 1 some may show 5, but they might publish more than what they show in your refresh interval (e.g. 1 week) In this case, using a stateless reader, you will start to miss article.

Yeah and of course it will be depend on your personality and risk model. Compared to other things I don’t want to risk my data, whether leaked or damaged. And I make mistakes, a lot. If you are very meticulous and can ensure that you can put up all the security measures yourself and won’t expose something you don’t want to. I am just not that kind of person.

I don’t think this is the right analogue. Having someone come to your door breaking things would take much larger effort, and easy to be caught. But DDoS or attack your service has minimal cost.

Visiting sites and sending the IP address is not the problem, the router has firewall and basically blocking unwanted attention. But when you expose something without protection and allow someone to burn your CPU, or, in a worse case, figure out your password for a not properly secured service, is a totally another issue.

I saw people setting up honey pot SSH and there are so many unauthorized access and I got scared. I think exposing entire machine to network is like you drive car without insurance. Sure you might be OK, but when trouble comes, it will be a lot of trouble.